Information processing system, information processing apparatus, information association method, learning method, and information association program

Abstract

An information processing system includes: an interface device that acquires medical action information related to a medical action for a user and consultation information related to a symptom of the user before and after the medical action; a storage that records the medical action information and the consultation information; and a processor that determines a treatment effect of the medical action for the user based on the consultation information acquired by the interface device and records the medical action information, the consultation information, and the treatment effect in association with each other in the storage.

Description

Technical Field

[0001] This invention relates to an information processing system, information processing device, information association method, learning method, and information association program that assist in treatment in medical institutions. Background Technology

[0002] In recent years, medical records and other information have been gradually digitized in medical institutions, making the management and use of patient medical records easier. For example, patients with illnesses may visit the hospital multiple times as needed for treatment. The records of surgeries, examinations, and other procedures performed on patients at the hospital are kept in their medical records. Electronic medical records facilitate retrieval and sharing of patient information. Doctors can refer to medical records during diagnosis to understand the progress of the patient's journey and provide more appropriate treatment.

[0003] As a method to confirm whether a doctor's treatment is appropriate, Japanese Patent Application Publication No. 2022-180692 (hereinafter referred to as Patent Document 1) proposes a technique for confirming the efficacy of a drug prescribed to a patient.

[0004] Existing technical documents

[0005] Patent documents

[0006] Patent Document 1: Japanese Patent Application Publication No. 2022-180692 Summary of the Invention

[0007] The problem the invention aims to solve

[0008] However, doctors may not be able to examine a patient until their symptoms are completely cured. For example, there are situations where patients stop seeking treatment or switch hospitals even if their symptoms only slightly improve or they lack the time to recover. Therefore, relying solely on hospital records sometimes makes it unclear why a patient stops seeking treatment or whether the hospital's treatment was effective. Furthermore, patients sometimes self-medicate before going to the hospital, making it difficult to determine the best treatment that comprehensively considers the patient's past and present condition.

[0009] However, on the other hand, it is difficult to get patients to input all of this type of information.

[0010] In view of this, the present invention aims to provide an information processing system, information processing apparatus, information association method, learning method, and information association program that can easily obtain insights related to effective medical actions appropriate to the patient's condition. Summary of the Invention

[0011] Solution for solving the problem

[0012] An information processing system according to one aspect of the present invention comprises: an interface device for acquiring medical behavior information related to a medical action for a user and consultation information related to the user's symptoms before and after the medical action; a memory for recording the medical behavior information and the consultation information; and a processor, wherein the processor performs the following processing: determining the treatment effect of the medical action for the user based on the consultation information acquired by the interface device; and recording the medical behavior information, the consultation information, and the treatment effect in a correlated manner in the memory.

[0013] An information processing apparatus according to one aspect of the present invention includes: an interface device comprising an information acquisition unit for acquiring medical behavior information related to medical behavior of a user; an information presentation unit for presenting information to a user; and a processor, wherein the processor performs the following processing: when the information acquisition unit acquires the medical behavior information, the processor generates a consultation form based on the medical behavior information for acquiring information related to the user's symptoms before and after the medical behavior of the user, and presents the consultation form on the information presentation unit.

[0014] One aspect of the present invention is an information association method that establishes a correlation between information on subjective symptoms and information on treatment effects. The information association method includes the following steps: a treatment result acquisition step, which obtains treatment results for a patient based on medical behavior information; a patient's subjective symptoms corresponding to the treatment up to the treatment; a treatment effect confirmation step, which confirms the treatment effect based on changes in the subjective symptoms after the treatment; and a step of establishing a correlation between the treatment result and changes in the patient's subjective symptoms before and after the treatment.

[0015] In one aspect of the learning method of the present invention, a neural network is learned using training data to construct an inference model, wherein the training data is obtained by annotating a dataset consisting of the time from the onset of initial symptoms before medical intervention until the medical intervention and at least one of the medical interventions, and the initial symptoms, up to the time when the individual feels cured, and by training the dataset.

[0016] One aspect of the present invention provides an information association program for establishing a correlation between information on subjective symptoms and information on treatment effects. The information association program enables a computer to perform the following steps: a treatment result acquisition step for obtaining treatment results for a patient based on medical behavior information; a patient's subjective symptoms confirmation step corresponding to the treatment up to the treatment being performed; a treatment effect confirmation step corresponding to the treatment and confirming the treatment effect based on changes in the subjective symptoms after the treatment; and a step for establishing a correlation between the treatment result and changes in the patient's subjective symptoms before and after the treatment.

[0017] The effects of the invention

[0018] According to the present invention, it has the effect of obtaining insights related to effective medical actions that are appropriate to the patient's condition. Attached Figure Description

[0019] Figure 1 This is a block diagram illustrating an information processing system according to one embodiment of the present invention.

[0020] Figure 2 This is an explanatory diagram showing an example of a message presented to the user through the control unit 11.

[0021] Figure 3 This is an explanatory diagram used to illustrate an example of knowledge information in external knowledge DB 19.

[0022] Figure 4 This is a flowchart used to illustrate the actions of the implementation method.

[0023] Figure 5 This is an explanatory diagram showing an example of a display of knowledge information from an external knowledge database 19, presented by the information presentation unit 17.

[0024] Figure 6 This is an explanatory diagram showing an example of an inference model produced by the Learning Department 19c. Detailed Implementation

[0025] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0026] (Implementation Method)

[0027] Figure 1This is a block diagram illustrating an information processing system according to one embodiment of the present invention. This embodiment is capable of acquiring information related to medical actions taken by a patient from the past to the present, such as medication, surgery, and treatment (hereinafter referred to as medical action information), and information related to the patient's symptoms and health (hereinafter referred to as consultation information). Based on this information, it can determine the effectiveness of the medical actions and present the determination result as a treatment outcome to doctors or the like. Thus, doctors or the like can obtain insights into effective medical actions corresponding to the patient's condition. Furthermore, in this embodiment, by utilizing a knowledge database based on the processing results, insights into effective medical actions related to the case can be obtained.

[0028] exist Figure 1 In this system, the portable terminal 10 can be operated by the patient themselves. Various information processing terminals, such as smartphones, tablets, and personal computers (hereinafter referred to as PCs), can be used as portable terminals. The portable terminal 10 includes a control unit 11, a sensor unit 12, an information input unit 13, a treatment result determination unit 14, a DB collaboration unit 15, an information processing unit 16, an information display unit 17, and an information recording unit 18. Each part of the portable terminal 10, such as the control unit 11, can be configured with a processor using a CPU (Central Processing Unit), GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), etc., and can be controlled according to a program stored in a memory (not shown). Some or all of the functions can also be implemented through hardware electronic circuits.

[0029] The control unit 11 accepts user operations for input operation units such as a keyboard and touchpad (not shown) provided on the portable terminal 10, and controls various parts of the portable terminal 10 based on the user operations. Furthermore, the control unit 11 controls various parts of the portable terminal 10 according to the application running on the portable terminal 10.

[0030] The sensor unit 12 and the information input unit 13 are interface devices for acquiring various types of information, constituting an information acquisition unit. The sensor unit 12 can be composed of one or more sensors, such as a camera or a microphone. The sensor unit 12 is controlled by the control unit 11 to acquire medical behavior information related to the user's medical behavior. For example, if the sensor unit 12 is composed of a camera, medical behavior information printed on medical expense reports, medical records, etc., can be acquired by photographing them with the sensor unit 12. In addition, medical expense reports include receipts issued by medical institutions (hospitals) where medical behavior has been performed, as well as receipts issued when purchasing over-the-counter drugs. The control unit 11 uses known image analysis techniques to identify strings recorded in medical expense reports, etc., based on the identification results, to acquire the user's own medical behavior information and record it in the information recording unit 18.

[0031] Medical behavior information refers to information related to a user's own medical behavior. For example, medical behavior information includes information about the medical institution the user visited, information about the treatment performed at the medical institution (treatment information), and information about prescriptions (prescription information). Furthermore, prescription information includes not only information about prescription drugs but also information about over-the-counter drugs, i.e., medication information related to the user's own medication use, including information such as the type of drug and the duration of use. Additionally, while nutritional supplements differ from pharmaceuticals, in this embodiment, the use of nutritional supplements may also be included in the medication information as needed. Moreover, even if medical behavior information is not explicitly recorded in the medical expense statement or similar documents, the control unit 11 can record the inferred medical behavior information in the information recording unit 18 if it can infer it from the information recorded in the medical expense statement or similar documents.

[0032] Thus, since medical behavior information is generated under various circumstances, it is sufficient to acquire and record it according to the specific situation at each time. For example, medical behavior information can be recorded in the information recording unit 18 of the portable terminal 10, or the recording units of the hospital's internal system (not shown), the pharmacy system, and the external knowledge database 19 shown can also have medical behavior information recording functions. In this case, an input unit can also be set up in the hospital system to send the medical behavior information input through the input unit to the portable terminal 10.

[0033] The information input unit 13 can be composed of input operation units such as a keyboard and touchpad, or it can be composed of various interfaces for receiving data from external sources. The information input unit 13 acquires consultation information and outputs it to the control unit 11 through user input operations and the reception of external data. For example, the information input unit 13 can also acquire consultation information through character input or voice input by the user. The control unit 11 provides the consultation information to the information recording unit 18 for recording. Similarly, it can collaborate with hospital and pharmacy systems to obtain medical behavior information.

[0034] The consultation information includes the user's medical history, symptoms, living environment, and past medical history. In this embodiment, it includes not only information obtained by doctors through consultation, but also information related to the characteristics and course of symptoms the user has experienced from the past to the present (self-reported symptoms). Furthermore, while not strictly self-reported symptoms, the consultation information includes the user's historical records of sleep, exercise, and other behaviors, as well as biological information such as body temperature, pulse, and blood pressure. The consultation information may also include information related to the user's living environment and hobbies (hereinafter referred to as lifestyle information). For example, lifestyle information related to the disease, such as drinking and smoking, can be obtained as consultation information. Since self-reported symptoms are sometimes not consciously perceived, the consultation content can be designed not only for the patient to write but also for the patient to choose from, allowing for measures to make the patient aware of their symptoms. Such consultations can be conducted using inference engines or other methods to infer consultation items, or the content of consultation items can be recorded in a database and selectively read and displayed.

[0035] Furthermore, an example is shown where the sensor unit 12 acquires medical behavior information and the information input unit 13 acquires consultation information. However, the sensor unit 12 can also acquire both medical behavior information and consultation information, and the information input unit 13 can also acquire both consultation information and medical behavior information. For example, the sensor unit 12 can also acquire biometric information such as pulse (consultation information). For example, if the user has a biometric information acquisition device such as a smartwatch, which can acquire biometric information, consultation information such as symptoms can also be acquired by that biometric information acquisition device. In addition, medical behavior information can also be acquired through input operations on the information input unit 13.

[0036] The information recording unit 18 comprises a recording device equipped with a recording medium (not shown) and has multiple areas for recording medical behavior information, consultation information, and daily life information (hereinafter referred to as medical information). In the example shown in Figure 18, the control unit 11 records medical behavior information in the medical behavior information recording area 18a, consultation information in the consultation information recording area 18b, and daily life information from the consultation information in the daily life information recording area 18c. Furthermore, the recording medium can also be located in an in-hospital system, pharmacy system, or external knowledge database 19, separate from the portable terminal 10. In such cases, a recording control unit corresponding to each recording medium can be installed, enabling each recording medium to cooperate with the terminal's information recording unit 18.

[0037] exist Figure 1 In the example, we know that the patient visited a doctor on November 3rd, received treatment A, and was prescribed prescription 1, experiencing symptoms of lower back pain and fatigue. Furthermore, we know that the patient had no symptoms two days prior to the visit, on November 1st, but developed lower back pain and fatigue on November 2nd, the day before the visit. Additionally, if the system is pre-programmed to know the user's drinking and smoking habits, this information can sometimes be used to improve diagnostic accuracy, predict complications, provide treatment recommendations, and remind the user to take prescription medications. Moreover, we know that on November 4th, the day after the visit, although fatigue persisted, the lower back pain disappeared, and by the day after the visit, November 5th, both symptoms had resolved, indicating a full recovery.

[0038] Thus, in this embodiment, information is acquired and recorded not only after the medical visit (after the medical procedure) but also before the medical visit (before the medical procedure). For example, information related to medical procedures such as over-the-counter medications taken by the user before the visit is recorded, as well as consultation information including the user's subjective symptoms, behavioral history, biological information, and lifestyle information from before the visit to the present. Furthermore, the information recording unit 18 also records file information including the user's name, age, gender, nationality, height, and weight through the control unit 11. This allows for the organization of knowledge information according to the patient's file.

[0039] The control unit 11 can also display messages for obtaining consultation information on a display device (not shown). For example, when the sensor unit 12 reads in medical expense details, the control unit 11 determines the contents of the medical expense details and creates a consultation form for obtaining consultation information corresponding to the determined medical behavior information. A message prompting the user to input information into the consultation form is displayed on the display screen of the display device. Furthermore, the consultation form is a table for obtaining the user's consultation information from the past to the present, showing the consultation items used to obtain the prescribed consultation information.

[0040] Figure 2 This is an explanatory diagram showing an example of a message presented by the control unit 11 to the user for inputting into a medical record form. Figure 2 In the example, the control unit 11 can also display a message such as "If you had a medical visit on November 3, please fill in the symptoms you have experienced up to the present, within the range that you can recall" as "Confirmation of symptoms until treatment", and the information input unit 13 can accept the user's input operation on their medical information from the past to the present.

[0041] External knowledge database 19 records knowledge information, including what symptoms are likely to occur in certain diseases (even if the disease name is unknown), what kind of examinations, tests, and treatments are needed, and what kind of prescriptions are required. This knowledge information can also be read into the information recording unit 18 of the portable terminal 10. Using this knowledge information, the user can retrieve or infer the symptoms based on the examination, test, and prescription information. If such inference is used, the control unit 11 can present candidate symptoms to the user, allowing the user to select a symptom. In other words, by utilizing the knowledge information, the control unit 11 avoids conducting inquiries about information irrelevant to the user.

[0042] In addition, after reading in medical expense details, the control unit 11 can display a message such as "You had a medical visit on November 3rd, please fill in your current symptoms" every day as a "confirmation of the treatment effect after treatment", and the information input unit 13 can accept the user's medical information input operation.

[0043] In this situation, since treatment and prescriptions are typically initiated due to concern about symptoms, the alleviated symptoms are the same as before the treatment. Therefore, the same information regarding the consultation can be used. Furthermore, by referring to past consultation results related to the user, it is possible to retrieve the types of symptoms present. Based on these results, a consultation can be conducted, for example, asking, "Has your back pain improved since the prescription?", allowing the user to answer with a yes or no. Additionally, by asking, "How much has your condition improved compared to before treatment?", numerical data can be obtained. Based on this simulated information, it is possible to determine how long it will take to recover, providing more accurate information until a full recovery, and acquiring data to present this information.

[0044] In this way, the user himself records information from the past, including before the medical visit (before the medical procedure), to the present, including after the medical visit (after the medical procedure), in the information recording section 18.

[0045] This example illustrates a post-medical consultation conducted after a consultation prior to the medical procedure. However, even if a consultation cannot be entered and recorded after a medical procedure, it is possible to conduct a consultation after the visit (following the medical procedure) or at the time when medication takes effect. Based on the results, consultation information can be obtained by recalling the circumstances at the time prior to the medical procedure. In this case, it is also possible to retrieve knowledge information and narrow down the scope of the consultation.

[0046] Furthermore, information regarding the user's own behavioral history and biometrics can be automatically acquired by the sensor unit 12 and recorded by the information recording unit 18. For example, information that can be automatically acquired by the sensor unit 12 includes the number of times the user went to the toilet, weight, blood oxygen saturation, blood pressure, number of exercise sessions, exercise frequency, number of outings, number of meals, number of coughs, and voice quality. Therefore, even without the user inputting any information, a certain level of diagnostic information can be obtained.

[0047] The treatment result determination unit 14 reads the medical behavior information and consultation information recorded in the information recording unit 18, and determines the treatment effect based on the read information. For example, the treatment result determination unit 14 can also determine whether the treatment is effective based on whether the patient's subjective symptoms have been cured from the date of treatment until the number of days determined based on the treatment information and prescription information. The treatment result determination unit 14 outputs the medical behavior information, consultation information, and treatment effect determination results read from the information recording unit 18 (hereinafter, this information is referred to as terminal medical information) to the DB collaboration unit 15 and the information processing unit 16. In addition, the terminal medical information may also include the user's profile information.

[0048] The processing result determination unit 14 performs learning based on a large amount of training data, such as deep learning. Deep learning is a multi-layered structured implementation of machine learning using neural networks. A representative example is the "feedforward neural network," which makes decisions by sending information forward. In its simplest form, this network only requires three layers: an input layer with m1 neurons, an intermediate layer with m2 neurons (with given parameters), and an output layer with m3 neurons corresponding to the number of classes to be classified. Furthermore, the neurons in the input layer and the intermediate layer, and the intermediate layer and the output layer, are connected by connection weights, and the intermediate and output layers are easily connected by applying bias values ​​to form logic gates. For simple discrimination, three layers may be used, but by using multiple intermediate layers, it is possible to learn combinations of multiple features during machine learning. In recent years, considering the relationship between learning time, judgment accuracy, and energy consumption, neural networks with 9 to 152 layers have reached a practical level. The network N1 used in machine learning can be any known network. For example, R-CNN (Regions with CNN), which utilizes a CNN, can also be used. Features: Regions based on CNN features, FCN (Fully Convolutional Networks), etc. This is accompanied by a process called "convolution" that compresses the feature amount of the image, operating with minimal processing and strong pattern recognition capabilities. Alternatively, "recurrent neural networks" (fully connected recurrent neural networks) can be used to process more complex information, support information analysis where meaning changes according to sequence and order, and allow bidirectional information flow.

[0049] After the DB Collaboration Unit 15 organizes the terminal medical information from the treatment outcome determination unit 14 into the information required by the external knowledge database (DB) 19, it outputs it to the external knowledge DB 19. In addition, the external knowledge DB 19 is a database that accumulates knowledge information, which is the result of establishing a correlation between past medical behavior information, consultation information, and treatment outcome determination results.

[0050] Doctors, experts, and scholars digitize and record their knowledge, various papers, and medical books in external knowledge database DB 19. That is, they organize and record the treatment of various diseases as knowledge information, but it is designed to allow retrieval or inference of the symptoms from the information of diagnosis, examination, and prescription based on the knowledge information.

[0051] In addition, external knowledge database DB 19 records knowledge information including the typical time from onset to medical intervention (although individual differences exist, an average tendency is used) and the typical time from medical intervention to recovery (although individual differences exist, an average tendency is used). Using this knowledge, doctors can communicate the approximate time of the next appointment to patients, and patients can focus on their treatment by anticipating their recovery time.

[0052] As described above, some or all of the knowledge information can also be recorded in the information recording section 18 of the portable terminal 10. The knowledge information recorded in the external knowledge database 19 is extremely comprehensive and general knowledge information, including gender, age, race, past medical conditions, etc. Since it is also an accumulation of past information, further information accumulation is needed regarding more individual conditions, diseases that have not occurred to date, and symptoms. With the development of IT technology and portable terminal technology in recent years, it is relatively easy to collect individual medical cases and achieve big data processing.

[0053] In this invention, treatment results corresponding to the diagnosis and treatment information of each patient are obtained from a portable terminal 10, etc., and the treatment effect is confirmed by checking the patient's subjective symptoms up to the treatment and by checking the changes in subjective symptoms after treatment. This establishes an information association between the treatment results and the changes in the patient's subjective symptoms before and after treatment. That is, in this embodiment, information of the same type as information provided through knowledge information is obtained, and knowledge information can be retrieved from the external knowledge database 19 or used as a reference to continuously update the external knowledge database 19.

[0054] Furthermore, as with the information recording unit 18, a large amount of data can be easily obtained as training data for the inference model by pre-processing the data. For example, it is possible to construct an inference model that obtains the output of consultation questions by inputting a medical expense statement, and it is also possible to replace the retrieval of external knowledge DB 19 by utilizing the inference results of the constructed inference model.

[0055] To achieve big data processing, a case record department can be set up in the external knowledge database 19 to record individual cases. Alternatively, individual cases can be collected and recorded in record departments set up within the hospital's internal system or other specialized service systems. This application envisions collaboration with other systems, enabling the collection and recording of big data through some form of communication with personal portable terminals 10 held by individuals. The collected big data can be utilized under necessary management and contractual conditions, and can be used to construct databases according to standardized specific rules, serving as training data for inference models for specific purposes.

[0056] In addition, external knowledge DB 19 can record not only raw data, but also data obtained by performing statistical processing on the raw data in a way that makes personal information and other details unknowable.

[0057] In addition, the external knowledge database 19 includes a control unit 19a, a training data digitization unit 19b, a learning unit 19c, and an inference unit 19d. The control unit 19d manages communication and recording in a manner that does not raise privacy issues, or properly organizes the database. The training data digitization unit 19b digitizes training data such as data recorded in the external knowledge database 19. The learning unit 19c uses the training data produced by the training data digitization unit 19d to learn and build an inference model. The inference unit 19d uses the inference model built by the learning unit 19c to output specific inference results for specific data inputs. The control unit 19a performs various controls, including data exchange between the training data digitization unit 19b, the learning unit 19c, and the inference unit 19d. The control unit 19a is also responsible for the following controls: responding to the portable terminal 10 with inference results upon request from the portable terminal 10, or retrieving specific data upon request from the portable terminal 10 and responding to the portable terminal 10 with the results.

[0058] Figure 6 This is an explanatory diagram showing an example of an inference model produced by the Learning Department 19c. In Figure 6 In the examples, the horizontal axis represents time, and the vertical axis represents the degree of disease progression. For instance, three examples of the same case illustrate the time required for recovery. Time t0 represents the time when initial symptoms appeared before medical intervention, time t1 represents the time when medical intervention was initiated in each example, and time t2 represents the time when the user felt cured in each example.

[0059] Figure 6 The lower part is an example of medical intervention performed relatively early, when initial symptoms appeared. Figure 6 The middle part refers to examples where medical intervention was performed relatively late, after the initial symptoms had appeared. Figure 6 The upper part is after the initial symptoms appeared. Figure 6 The lower part and Figure 6 An example of a medical procedure performed at the right time between the middle and lower parts of the body. Figure 6 In the example, it was argued that the timing of medical intervention after the onset of initial symptoms is related to the time required for a cure.

[0060] The training data processing unit 19b annotates the time (t2-t1) obtained through consultation until the individual feels cured, based on a dataset containing three types of data: information on initial symptoms prior to medical intervention obtained through consultation, information on the time from the onset of initial symptoms to the medical intervention (t1-t0), and information on the content of the medical intervention. Furthermore, age, gender, ethnicity, etc., can be added to the aforementioned dataset when creating the training data.

[0061] The learning unit 19c uses training data produced by the training data processing unit 19b to learn and construct an inference model. Based on this inference model, the initial symptoms and the time from the appearance of the initial symptoms to the intervention of medical procedures are used as inputs to obtain the time required for recovery, which is then used as the inference result. This inference result can be displayed on the portable terminal 10.

[0062] Furthermore, specific information about medical actions can be obtained through consultation and incorporated into the training data. Therefore, the training data during learning also includes information related to medical actions such as the time from the onset of symptoms to medical action, purchasing and taking over-the-counter medications at a pharmacy, and visiting a clinic and taking prescribed medications. Thus, the inference model can output predicted healing times for cases where medical action is taken immediately after onset, two or three days later, or more than a week later. By enabling this information to be sent to and displayed on the portable terminal 10, users can obtain extremely useful information related to recovery.

[0063] Furthermore, if the external knowledge database 19 exists on a network such as the Internet, the database collaboration unit 15 uses a communication function (not shown) of the portable terminal 10 to send terminal medical information to the external knowledge database 19. Additionally, the database collaboration unit 15 and the external knowledge database 19 exchange information, outputting knowledge information related to the specified case to the information processing unit 16. As described above, the portable terminal 10 can also have the same functions as the external knowledge database 19 by sending a portion of the knowledge information from the external knowledge database 19 to the portable terminal 10 and recording the knowledge information in the information recording unit 18 of the portable terminal 10.

[0064] Figure 3 This is an explanatory diagram used to illustrate an example of knowledge information in external knowledge DB 19.

[0065] External knowledge database DB 19 has multiple recording areas R1, R2, ... (hereinafter, these areas are collectively referred to as recording areas R) for recording knowledge information (related information) for each symptom. Figure 1 For example, knowledge information related to abdominal pain is recorded in recording area R1, and knowledge information related to fever is recorded in recording area R2. Each recording area R has multiple disease-specific recording areas RD for recording knowledge information for each disease.

[0066] Knowledge information for each candidate disease is recorded in the disease-specific recording area RD. A candidate disease is a potential disease that presents with the given symptom; multiple candidate diseases are considered for a single symptom. For example, even with fever, multiple disease candidates, such as those caused by COVID-19 or influenza, are considered. The disease-specific recording area RD may also record information related to representative symptoms, causes, examination results, in-hospital treatment, and prescribed medications for that disease, serving as knowledge information.

[0067] Furthermore, in this embodiment, external knowledge database 19 is provided with terminal medical information by the database collaboration unit 15. Figure 1 In the example, terminal medical information is recorded as archival information in the disease-specific recording area RD of candidate disease 1. In this archival information, terminal medical information from DB Collaboration Department 15 is recorded as information associated with existing knowledge information. External Knowledge DB 19 can, for example, determine which disease-specific recording area RD to associate the terminal medical information with based on a comparison between the subjective symptoms contained in the terminal medical information from DB Collaboration Department 15 and representative symptoms in the disease-specific recording areas RD within the external knowledge DB 19. Furthermore, external knowledge DB 19 can also determine which disease-specific recording area RD to associate the terminal medical information with based on a comparison between the medical behavior information contained in the terminal medical information and in-hospital treatments and prescription drugs in the disease-specific recording areas RD within the external knowledge DB 19. Figure 1 In the example, as information that established the association, the following information was recorded: as a subjective symptom, it was abdominal pain and throbbing pain, prescription 1 was issued, and the patient recovered in two days.

[0068] Alternatively, as a method of association, an inference model utilizing deep learning can be used, similar to the processing result determination unit 14.

[0069] After processing the terminal medical information from the treatment result determination unit 14 or the knowledge information from the DB collaboration unit 15 as required for presentation, the information processing unit 16 outputs it to the information presentation unit 17. The information presentation unit 17 may be composed of, for example, a display device (not shown) and a speaker installed on the portable terminal 10, and can present the terminal medical information from the information processing unit 16 by displaying it on the display screen of the display device or by voice output. Furthermore, the information presentation unit 17 can also be controlled by the control unit 11 to present a consultation form created by the control unit 11 to the user. Additionally, the portable terminal 10 may have a communication function (not shown) that enables connection to a medical institution, and the information presentation unit 17 can use this communication function to send the terminal medical information from the information processing unit 16 to the medical institution (or doctor). Through the presentation by the information presentation unit 17, the user and the medical institution (or doctor) can determine the extent to which the symptoms have improved through the user's or the medical institution's treatment.

[0070] Furthermore, the information presentation unit 17 can present knowledge information from the external knowledge database 19 to the user or medical institution. This allows the user or medical institution to easily determine which treatment is effective for their current symptoms. For example, the control unit 11 can also control the database collaboration unit 15, the information processing unit 16, and the information presentation unit 17 to present the knowledge information retrieved from the external knowledge database 19 to the user or medical institution (doctor) even if the treatment effect determination unit 14 has not obtained a determination result, or based on the user's own actions. For example, the control unit 11 can also retrieve knowledge information from the external knowledge database 19 based on the user's medical actions or symptom input, extract and present the knowledge information as the search results.

[0071] Next, refer to Figure 4 and Figure 5 The operation of an implementation method configured in this way will be explained. Figure 4 This is a flowchart illustrating the actions of the implementation method. Additionally, Figure 5 This is an explanatory diagram showing an example of a display of knowledge information presented by the information presentation unit 17 using knowledge information from external knowledge DB19.

[0072] exist Figure 4In step S1, the control unit 11 determines whether there is an input operation to release the portable terminal 10 from sleep mode. S1 is a standby state for releasing sleep mode requests. When a sleep mode release request is generated (S1 "Yes"), the control unit 11 displays a predetermined home screen on the display screen of a display device (not shown) (S2). In S3, the control unit 11 determines whether a request has been made to launch an application for creating and utilizing terminal medical information (medical application).

[0073] When a request to launch a medical application is generated in S3 (S3 is "Yes"), the control unit 11 determines whether there is an input operation such as a medical expense statement. If there is an input operation such as a medical expense statement (S4 is "Yes"), the control unit 11 obtains medical behavior information in S5 (S5), and displays the message required for consultation based on the medical expense statement information and obtains consultation information based on the retrieval and inference of the knowledge database (S6).

[0074] Consultation information is accessed through user reference. Figure 2 Input can be obtained through a display as shown (or voice communication), but the scope can be narrowed down based on the medical expense details, thereby simplifying the problem. User input methods can also include button operation, text input, character recognition, voice recognition, other touch input, or Yes / No selection input using a vibration sensor. Specifically, the processor of the portable terminal 10 is characterized in that, through these input / output controls, it performs the following processing: when the information acquisition unit acquires medical behavior information related to the user's medical behavior, it determines the medical behavior information, generates consultation information to acquire information related to the user's symptoms before and after the aforementioned medical behavior, and presents it on the information display unit.

[0075] In addition, when the processor obtains information related to the symptoms of the user prior to the medical behavior based on the obtained medical behavior information, it selects candidate symptoms and creates the consultation information. When it obtains information related to the symptoms of the user after the medical behavior based on the input information, it selects candidate symptoms and creates the consultation information.

[0076] Furthermore, when generating consultation information, information such as the initial symptoms prior to the medical intervention, learned from previous consultations, and the time from the onset of those initial symptoms to the conclusion of the medical intervention are also obtained. Additionally, in consultations following a medical intervention, the time from the date and time of the medical intervention (examination, medication) until the patient feels cured can also be obtained. Consultation information in this case can also be obtained by inputting it through the user information input unit 13 in the form of voice, text, or question-and-answer format. In the absence of user input, information obtained based on biometrics can be used as a substitute, and the consultation information in this case can be calculated through computation. If it is after a medical intervention, consultation information in this case can also be obtained by repeatedly checking the patient's physical condition daily.

[0077] The portable terminal 10 has a recording unit (database recording unit) in the DB collaboration unit 15 or the information recording unit 18 for recording knowledge information. When obtaining medical behavior information related to medical behavior, it can collaborate with the external knowledge DB 19 and the records in the information recording unit 18 to determine the candidate symptom examples for consultation before the aforementioned medical behavior. Furthermore, collaboration refers to retrieval and inference. The external knowledge DB 19, which collaborates with the DB collaboration unit 15, records information about the relationship between the symptoms of each disease, medical behavior, and the effects of that medical behavior—that is, knowledge information. Alternatively, information filled out by a doctor or pharmacist can be used instead of medical behavior information.

[0078] The control unit 11 records the acquired medical behavior information and consultation information in the information recording unit 18 (S7) and returns the processing to S2. In addition, if there is no request to start the medical application ("No" in S3), the control unit 11 performs other functions of the portable terminal 10 in S18 and returns the processing to S2.

[0079] If the control unit 11 determines in S4 that no input operation such as the medical expense details statement has been performed, the process is transferred to S10 to determine whether an information presentation mode has been indicated. If no information presentation mode has been indicated (S10 "No"), the control unit 11 returns the process to S2. If an information presentation mode has been indicated (S10 "Yes"), the control unit 11 transfers the process to S11 to determine the type of information to be presented. If terminal medical information has been indicated as the information to be presented, the control unit 11 controls the information processing unit 16 and the information presentation unit 17 to display the medical behavior information and consultation information read from the information recording unit 18, as well as the terminal medical information including the treatment effect determination result from the treatment result determination unit 14, on the display screen of the display device or send it to the medical institution or doctor (S12). In this way, the user or the medical institution (or doctor) can judge the validity of the user's medical behavior from the past to the present by referring to the presented terminal medical information.

[0080] Furthermore, when knowledge information is specified as the information to be presented, the control unit 11 displays a symptom input message to determine the knowledge information to be retrieved, and acquires the symptom input provided by the user (S13). The control unit 11 controls the DB collaboration unit 15 to send the symptom input result to the external knowledge DB 19 and acquire the knowledge information corresponding to the symptom. The external knowledge DB 19 searches for people (disease candidates) corresponding to the input symptom and sends the knowledge information as the search result to the DB collaboration unit 15. In addition, if the external knowledge DB 19 does not obtain the corresponding search result, it sends the information indicating the meaning to the DB collaboration unit 15.

[0081] If the control unit 11 obtains a search result from the external knowledge database 19, meaning it can find a person (disease candidate) with a similar case corresponding to the user's symptom input in the external knowledge database 19 (S14 "Yes"), it controls the information processing unit 16 and the information presentation unit 17 to display knowledge information indicating the course of the retrieved person's illness (S15), and returns the processing to S13. If the control unit 11 does not obtain a search result corresponding to the user's symptom input (S14 "No"), it transfers the processing to S16, indicating that no information was obtained.

[0082] Furthermore, several methods are considered for displaying the knowledge information in S15. First, for a specific disease, a member managing the external knowledge database 19 selects typical patient examples and sends the selected information to the portable terminal 10, enabling it to be displayed as described later. Figure 5 That's how it will be displayed.

[0083] Furthermore, when a large amount of patient information (dataset) is collected, the following methods exist: A graph is created with the horizontal axis representing the time taken to cure and the vertical axis representing the corresponding number of cases; the average cure time is calculated; or the peak value on the vertical axis is used to calculate the cure time for the largest number of cases, and this information is sent to the portable terminal 10 for display. Alternatively, the system can retrieve examples of patients cured earlier than the average cure time and examples of patients whose cure time deviated from the average cure time by 2 or 3 sigma through statistical calculations, and send these to the portable terminal 10 for display. Thus, users can obtain reference information on what actions might worsen the condition and what actions might lead to an earlier cure.

[0084] Alternatively, the output of the inference model learned using external knowledge DB 19 can be acquired and displayed. Furthermore, if the portable terminal 10 has a system architecture with a built-in inference model, the output based on that inference model can also be displayed.

[0085] As described above, this inference model is learned using training data. This training data is obtained by analyzing how long it takes to recover after a medical procedure (whether the symptoms before the medical procedure were alleviated) based on imagined medical behavior annotations such as subjective symptoms, detailed medical expense records, and pharmacy purchase history. Therefore, such an inference model can be used to infer what will happen after receiving a particular medical procedure. Through this inference, users can choose which medical procedure is preferred and execute it themselves. That is, when symptoms are difficult to predict, it is possible to determine whether to rest at home or go to the hospital based on the experience of other users. Furthermore, the recovery time varies greatly from person to person due to factors such as physical strength; therefore, when learning the inference model, it can be configured to accept individual profile information for more accurate inferences. Additionally, for example, even if similar subjective symptoms, such as "fever," are measured using medical devices, the inference will vary depending on the disease, such as "cold," "COVID," or "flu." This can be determined by inputting information from detailed medical expense records and prescriptions to differentiate medical behaviors. Therefore, when "fever" is input as a subjective symptom, the model can also be configured to output information about these different diseases. In addition, it is possible to determine which of these diseases is likely based on the patient's behavioral information. For example, if the patient's place of residence is entered into the file, the system can help display the most probable disease names based on details of medical expenses incurred in that area during that period.

[0086] Figure 5 An example of the knowledge information displayed in S15 is shown.

[0087] Figure 5 The example shown is a display of knowledge information Dk displayed on a display screen 17a of a display device (not shown) provided on a portable terminal 10. Figure 5 The knowledge information displayed in Dk shows the search results for three disease candidates, patterns A through C, obtained from external knowledge database DB 19. Pattern A indicates that for someone who went to the hospital, took medication, and took leave for recuperation, the disease has been cured. Pattern B indicates that for someone who did not go to the hospital but only took medication and continued working, the period until recovery was prolonged. Pattern C indicates that for someone who neither went to the hospital nor took medication, the disease has worsened.

[0088] The inference model of external knowledge DB 19 is learned using training data. This training data is obtained by considering how long it takes to cure (whether the subjective symptoms before the medical treatment are relieved) based on the annotations of medical behaviors imagined based on subjective symptoms, medical expense details, and purchase history information in pharmacies. Therefore, it can infer the effect of receiving a certain medical behavior based on the subjective symptoms input by the person.

[0089] Users can refer to the knowledge information displayed in Dk, and use the results of medical actions and their effectiveness on people with the same symptoms as themselves as a reference for their own medical actions. Furthermore, medical institutions or doctors can also access this knowledge information, which can provide effective assistance in their treatment of patients.

[0090] Thus, in this embodiment, information on past medical procedures and treatments can be acquired, along with corresponding consultation information. Based on this medical procedure information and consultation information, a determination of the treatment effectiveness is made, and the terminal medical information, including this medical procedure information, consultation information, and the determination of the treatment effectiveness, is presented in a linked manner. Therefore, the effectiveness of medical procedures can be determined based on the presented terminal medical information, and assistance can be provided for future medical procedures.

[0091] This invention is not limited to the embodiments described above. During implementation, the constituent elements can be modified and embodied without departing from its spirit. Furthermore, various inventions can be formed through appropriate combinations of the multiple constituent elements disclosed in the above embodiments. For example, several constituent elements shown in the embodiments may be deleted. Moreover, the constituent elements involved in different embodiments may be appropriately combined.

[0092] Furthermore, in the technology described herein, most of the controls and functions illustrated primarily by flowcharts can be set through a program, and the aforementioned controls and functions can be achieved by a computer reading and executing the program. This program can be recorded or stored, in whole or in part, as a computer program product on removable media such as floppy disks, CD-ROMs, or non-volatile memory, or on storage media such as hard disks or volatile memory, and can be distributed or provided at the time of product shipment or via removable media or communication lines. Users can easily implement the information processing system, information processing apparatus, information association method, learning method, and information association program of this embodiment by downloading the program via a communication network and installing it on a computer, or by installing it from a recording medium on a computer.

Claims

1. An information processing system, comprising: An interface device that acquires medical behavior information related to medical behavior of a user and consultation information related to the user's symptoms before and after the medical behavior. A memory that records the medical behavior information and the consultation information; as well as processor, The processor performs the following processing: Based on the consultation information obtained by the interface device, the effectiveness of the treatment for the user's medical behavior is determined; and The medical behavior information, the consultation information, and the treatment effect are recorded in the memory in an associated manner.

2. The information processing system according to claim 1, wherein, The processor determines the treatment effect of the medical procedure based on a comparison between the first symptom and the second symptom, wherein the first symptom is the user's symptom before the medical procedure, and the second symptom is the user's symptom after the medical procedure.

3. The information processing system according to claim 1, wherein, The processor performs the following processing: The medical behavior information obtained by the interface device is compared with a database containing accumulated knowledge information to extract candidate symptoms. The knowledge information is the result of establishing a correlation between medical behavior information related to medical behavior, consultation information related to symptoms, and the judgment results of the treatment effect of medical behavior. The interface device enables the user to select from the extracted candidate symptoms; and The symptoms selected by the user are used as the consultation information.

4. The information processing system according to claim 1, wherein, The processor obtains the medical behavior information based on the user's medical expense details or medical records.

5. The information processing system according to claim 1, wherein, The processor obtains the medical procedure information from the hospital where the medical procedure was performed.

6. The information processing system according to claim 1, wherein, When the interface device consists of a device that accepts the user's character input and voice input, the processor obtains the consultation information based on the user's character input or voice input. When the interface device is composed of a biometric information acquisition device owned by the user, the processor acquires the consultation information based on the output of the biometric information acquisition device.

7. The information processing system according to claim 1, wherein, In addition to acquiring the symptoms, the processor also acquires the user's lifestyle information as the consultation information.

8. The information processing system according to claim 1, wherein, The processor compares the information with a database containing accumulated knowledge information to extract and present knowledge information similar to the medical behavior or symptoms of the user. The knowledge information is the result of establishing a correlation between medical behavior information related to the medical behavior, consultation information related to the symptoms, and the judgment results of the treatment effect of the medical behavior.

9. The information processing system according to claim 8, wherein, If no determination of the treatment effect is obtained, the processor extracts knowledge information from the database that is similar to the medical behavior or symptoms of the user and presents it.

10. An information processing device comprising: An interface device, which has an information acquisition unit that acquires medical behavior information related to the user's medical behavior; The information presentation department, which presents information to users; and processor, in, The processor performs the following processing: when the information acquisition unit acquires the medical behavior information, the processor creates a consultation form based on the medical behavior information to acquire information related to the user's symptoms before and after the medical behavior, and presents the consultation form to the information presentation unit.

11. The information processing apparatus according to claim 10, wherein, When the processor obtains information related to the user's symptoms prior to the medical behavior based on the acquired medical behavior information, it selects candidate symptoms to determine the consultation items used to obtain the consultation information.

12. The information processing apparatus according to claim 11, wherein, When the processor obtains information related to the user's symptoms after the medical procedure based on the information obtained related to the user's symptoms before the medical procedure, it selects candidate symptoms to determine the consultation items used to obtain the consultation information.

13. The information processing apparatus according to claim 10, wherein, The processor sets up a consultation item for information related to the user's symptoms prior to the medical treatment. The consultation item is used to obtain the time from the time the symptoms appeared until the medical treatment, and / or the time relative to the date and time of the medical treatment until the user felt cured.

14. The information processing apparatus according to claim 10, wherein, It also includes a database collaboration unit that collaborates with a database containing accumulated knowledge information. This knowledge information is the result of establishing associations between medical behavior information related to medical actions, consultation information related to symptoms, and the assessment results of the treatment effects of medical actions. The processor performs the following processing: when it obtains medical behavior information related to the medical behavior of the user, the processor collaborates with the database to determine the symptoms that precede the medical behavior as candidates.

15. The information processing apparatus according to claim 14, wherein, The database records information about the relationship between the symptoms of each disease, medical procedures, and the effects of those medical procedures.

16. An information association method for linking information about subjective symptoms with information about treatment effects, the information association method comprising the following steps: The steps for obtaining treatment results for patients based on medical behavior information; The patient's subjective symptom confirmation steps corresponding to the treatment up to the treatment; A treatment effect confirmation step corresponding to the treatment, which confirms the treatment effect based on changes in the subjective symptoms after the treatment; as well as The step of establishing a correlation between the treatment outcome and changes in the patient's subjective symptoms before and after the treatment.

17. The information association method according to claim 16, wherein, In the patient's subjective symptom confirmation step, a database that records information on the relationship between symptoms, medical actions, and the effects of those medical actions is used to retrieve candidate symptoms that are associated with the medical actions.

18. The information association method according to claim 16, wherein, Regarding the steps for confirming subjective symptoms and confirming treatment effects, one of them should be confirmed first, and the result should be used to confirm the other.

19. A learning method, wherein, A neural network is learned using training data to construct an inference model, wherein the training data is obtained by annotating a dataset consisting of the time from the onset of initial symptoms before medical intervention until the medical intervention, and at least one of the medical interventions, and the initial symptoms, up to the time when the individual feels cured, and then digitizing the training data.

20. An information association program for establishing a correlation between information about subjective symptoms and information about treatment effects, said information association program for causing a computer to perform the following steps: The steps for obtaining treatment results for patients based on medical behavior information; The patient's subjective symptom confirmation steps corresponding to the treatment up to the treatment; A treatment effect confirmation step corresponding to the treatment, which confirms the treatment effect based on changes in the subjective symptoms after the treatment; as well as The step of establishing a correlation between the treatment outcome and changes in the patient's subjective symptoms before and after the treatment.

Images